Artificial Life Algorithm for Management of Multi-reservoir River Systems

The design and operation of civil engineering systems, particularly water resources systems, has been pursued from the perspective of minimizing costs and related negative impacts, maximizing benefits, or a combination thereof. Due to the complex, nonlinear nature of the majority of systems, together with an increase in digital computing capabilities, global search algorithms are becoming a common means of meeting these objectives. This paper employs an artificial life algorithm, derived from the artificial life paradigm. The algorithm is evaluated using standard optimization test functions and is subsequently applied to determine optimal dam operations in multi-reservoir river systems. The optimal dam operation scheme is that which indirectly minimizes environmental impacts caused by short-term water level fluctuations. Optimal releases are sought by coupling an artificial life algorithm with FLDWAV, a one-dimensional, steady flow simulation model. The resulting multi-reservoir management model is successfully applied to a portion of the Illinois River Waterway.